Custom motor sound profile tutorial

Ryan Douglas

Administrator
Staff member
As of RealFlight 6.00.044 (public beta) released on April 25, 2012, RealFlight allows for the use of custom motor sound profiles. This guide will explain the basics of recording, editing, and importing a custom motor sound profile.


Setting up your recording environment
Make sure you have the necessary hardware:

You'll want to use a good microphone to capture your motor sounds. A good microphone is one with good frequency response--one that equally captures all the high and low sound frequencies your motor makes. There are more aspects of microphone quality to consider, but that goes outside the scope of this guide. Consider a microphone with a wind screen to avoid picking up unwanted noise both from wind and from airflow generated by the aircraft.

It is a good idea to use a mic stand to mount the microphone during the recording process. This makes it easier to position and orient the microphone precisely where you want it. Avoid fixing the microphone directly to the motor or aircraft's fuselage, as this will introduce a lot of unwanted vibration noise into the recording. Also avoid holding the microphone by hand during recording, as this can introduce similar vibration/rubbing noises into the recording, as well as change the sound characteristics of the recording as the microphone moves around slightly.

You'll need a way to record the sounds in a digital format, either directly onto your computer, or onto a portable device so they can be transferred to your computer afterwards.

A tachometer is useful for measuring the exact RPM of your motor during the recording process.

You'll need to find a good location to record your motor. As much as possible, you must isolate the motor sound from all other noise. Recording outdoors often means you're fighting with the sounds of wind, wildlife, automobiles, full-scale airplanes flying overhead, etc. You may have a successful recording session if you can find a quiet outdoor area with little to no wind and minimal background noise. Recording indoors usually offers less background noise, but it also tends to carry the problem of short echoes coming from sound waves bouncing off the walls. This can result in resonant frequencies lingering in your audio recording, depending on the size of the room in which you're recording. Ideally, recording is best done in a professional studio with proper sound damping material mounted on the walls to absorb the echo. If you're on a budget, you can try replicating this setup at home by mounting unwrapped sleeping bags on the walls of a small room.

Note that a skilled digital audio producer can alleviate some of the echo and noise problems after the recording process is done, but it's generally much easier and less time-consuming to avoid these problems in the first place.


Recording your motor
You'll need to record your motor running at various RPMs--the highest it is capable of, the lowest, and several points in between. We recommend creating a single long recording of your motor stairstepping from its lowest RPM to highest (or high to low). The RPM should remain unchanged except when transitioning to the next value you want to record. The goal is not to record those transitions between RPM values; the sim will handle that automatically. The goal is to capture steady examples of discrete RPM values throughout the motor’s operating range.

There are many things to keep in mind for this process:
  • Clipping occurs when the sound is louder than the microphone is able to record, resulting in distortion. Your recording setup should have a microphone input sensitivity control. Set it to the highest level you can find that does not allow clipping. You should be able to record your sound source at its loudest moments without the levels going into the red (if you have meters) or otherwise triggering a clipping indicator. Reduce the sensitivity as needed until you can accomplish this.
  • Do not adjust the microphone sensitivity during a single take. Motors are louder at full throttle than at idle. Let the lower RPM values be naturally quieter in your recording, as well.
  • Put the microphone close to the motor. This will help you maximize the proportion of sound you get from the motor versus unwanted echoes and background noise. However, placing the microphone too close will result in the "proximity effect", where lower-frequency sounds are exaggerated. It may also result in clipping, which should always be avoided. Try different distances to see what sounds best.
  • Try recording from different angles as well. Recording from above the sound source may sound noticeably different than from the side. Again, experiment with microphone placement to find what sounds best.
  • A rule of thumb for determining the spacing of different RPM values to record is that each RPM should be about 25% higher than the one below it. For example, if you record a steady 4500 RPM as one of your “stair steps”, we recommend trying to record the next one at around 5625 RPM. If you have a musical ear, this translates to about 4 semitones, or a major third. Again, this is just a rule of thumb and does not need to be exact. You can record with denser or sparser gaps between RPMs if you like. There is no point getting values too close to each other, however. Allow some room between the samples or it might actually sound worse in the end.
  • Record each steady RPM for at least 10 seconds. The RPM should not vary during this time. The longer you record, the easier it will be for you to find a suitable section for looping later.
  • If you are recording a battery-powered electric motor, the maximum attainable RPM will drop almost immediately as the battery begins losing its charge. For this reason, you should start your recording with a fully charged battery, and you should record the highest RPM first. Note that even with this approach, the maximum RPM is likely to drop slightly while you attempt to record it. You can try to fix this later during the editing process, or you can record an RPM just barely less than the motor's maximum to avoid the slow drop in RPM.
  • If you have a tachometer, remember to measure the exact RPM of the motor for each sample. It may be easiest to vocally announce the RPM after each sample in the audio recording, and then refer to that later during the editing process.
  • Avoid recording to a compressed audio format such as MP3. This will help preserve the quality of the recording during the editing process.

Editing your recording
Once you've transferred your recorded audio to your computer, you'll need to edit your recording and prepare individual RPM samples for importing into RealFlight. This requires audio editing software. Audacity is a free audio editor and is available for download at http://audacity.sourceforge.net/ There are numerous commercial audio editors available as well, with varying levels of features.

The overall goal of the editing process is to separate your one long recording into several individual looped samples--one for each RPM you recorded. The individual samples can be saved either in .wav format or in .ogg (Ogg Vorbis) format, but all samples for a single sound profile must be saved in the same format. Ogg Vorbis audio files will have negligibly lower sound quality but will be much smaller in file size. That means dramatically smaller .RFX file sizes, so we recommend using the .ogg format.

Each individual sample should have a filename that matches the RPM it represents. For instance, a sample of a motor running at 4250 RPM should be named 4250.ogg or 4250.wav. Only numbers are allowed in the filename.

There are a handful of other things to keep in mind when editing each individual RPM sample:
  • Each sample should be around 7 seconds in length.
  • Make sure the sample sounds good when looped on repeat. There is often an audible "pop" when looping between the end of a sample and the beginning due to slight phase differences. This problem can generally be eliminated through trial and error by repeatedly shaving a few milliseconds of audio from the beginning and end of the sample until the loop sounds smooth.
  • Listen to the sample on repeat many times for artifacts and unwanted noises. Sometimes it can take a while to find an annoying subtle "squeak" or "thud" that made its way into your recording, but once you notice it, you'll hear it constantly. These artifacts can be eliminated with the use of equalizers and other editing techniques not covered by this guide.
  • Be sure the RPM in the recording matches the name you give to the file. RealFlight crossfades between different samples at intermediary RPMs, meaning two samples are often playing simultaneously. If one of them is assigned an accurate RPM but the other is not, you will hear the very dissonant sound of two slightly different tones being played together instead of blending properly. Advanced users can generate a reference tone in the audio editing software to check against. For example, if you expect your sample is 9000 RPM, you can divide 9000 RPM by 60 seconds to get 150 Hz. Generate a 150 Hz tone in the audio editor and play it alongside your 9000 RPM sample. If the two sounds played in tandem sound dissonant, or make noticeably different tones, your sample is probably not actually 9000 RPM.

Importing your sound profile into RealFlight
Once all your individual edited samples have been saved to the same folder, in the same file format, and with proper filenames, you can import them into RealFlight and create a new sound profile. Do this by clicking the "Simulation" menu, then "Import", and then "Motor Sound Profile...". The resulting import wizard will guide you through the import process.
 

willsonman

Well-known member
Ryan, feel free to delete this post but I think to tag on here we might post on some microphones that would be suitable for the average end-user. Its been several years (8 or so) since I was in the market for a nice microphone so I am having to do some homework for a good one. Any others out there with knowledge on this subject?
 

Madratter

New member
The truth is that for this kind of application, extremely high fidelity just isn't necessary. I use a Audio-Technica ATM31a in my home studio but that is no longer available. But I paid roughly 100$ for it. If anything it is way overkill for an application like this. It is just your basic Condenser mike. Any decent brand name mic in that price range is going to do way more than well enough for this sort of work.
 

Adam Taylor

Adam-inistrator
The Zoom H1 is indeed a handy little device with decent X/Y stereo microphones built-in. We've used it here at Knife Edge to record some of the sound profiles that wound up in RealFlight.

I'm also personally a big fan of the Shure SM57.
 

brentg

Well-known member
When saving in the ogg vorbis format what mono bitrate do you use? What sample rate for recording ? And could you use "Resampling" to manipulate original rpm sample/s?
 
Last edited:

Adam Taylor

Adam-inistrator
When saving in the ogg vorbis format what mono bitrate do you use? What sample rate for recording ? And could you use "Resampling" to manipulate original rpm sample/s?
I generally record to .wav with a bitrate of 24 and a sample rate of 96Khz, just because I can. When compressing to Ogg Vorbis later, however, the only important thing is to not lose too much of the original quality of the sound. If you play the compressed and uncompressed audio side-by-side and you can hear an audible difference, you've compressed too far.

You can certainly resample the original recorded samples if you want. However, if the goal is to create a sample with an RPM higher what you were able to record, there probably isn't much point. RealFlight automatically up-samples the highest available RPM sample to match the motor's current RPM.
 

Ryan Douglas

Administrator
Staff member
Something I meant to point out earlier: If you record an engine up to, say, 7600 RPM, and you (or someone else!) uses that sound profile for an engine that revs to 12000 RPM, it's going to sound bad. That's because all RealFlight can do in that situation is take the highest available RPM sample and adjust its sample rate as much as necessary, which is much farther than is reasonable in this example.

Though less likely to be an issue, the same potential problem exists at the low end. If your lowest recording of an electric motor is 3500 RPM, it's probably going to sound bad spinning at 1500 RPM in the sim.

This is just something to be aware of when choosing a sound profile for a particular application.
 

gibson

New member
Is it possible to use/record a REAL aircraft engine at its various RPMs, and then assign the recordings to a proportionate RPM in real flight? For example, if I want to record a real Continental Radial engine idling at about 600 rpm, and then at full power, approx. 1900 rpm, and have my 25% increments in between, would it then be possible to edit and upload it to real flight so as to create a sound profile that sounds like the real aircraft?
 

jeffpn

Well-known member
I imagine you could fart into a microphone at various RPMs and have it sound like a Real Fart in RF.
 

Madratter

New member
Is it possible to use/record a REAL aircraft engine at its various RPMs, and then assign the recordings to a proportionate RPM in real flight? For example, if I want to record a real Continental Radial engine idling at about 600 rpm, and then at full power, approx. 1900 rpm, and have my 25% increments in between, would it then be possible to edit and upload it to real flight so as to create a sound profile that sounds like the real aircraft?
Yes, but only in the range of RPMs that you record. That will be limiting since the engine actually used in modeling the aircraft in Realflight almost certainly goes way outside the RPM range you mention on the high end.
 

gibson

New member
So let's say that I recorded a real engine at 1900 rpm, is there a way to upload it and have that file be used by real flight as the max rpm sound for the engine?
 

Boof69

Well-known member
You can include it as 1900.ogg in the folder with the other recorded sounds. Even though it mos likely wouldn't sound right.
 

Ryan Douglas

Administrator
Staff member
If the 1900 RPM sample is the highest one you provide, then RealFlight will use it for some RPMs below 1900 (how many depends on the next lowest sample you provide) and for all RPMs above 1900. Making it do that doesn't require any special steps; it's simply the way the system works.

The problem is that the result will most likely be undesirable. Unless the engine using your custom sound profile also tops out around 1900 RPM, it's going to sound horrible. For example, imagine an engine that reaches 10,000 RPM that is using your custom sound profile that only goes to 1900. When it's operating at, say, 9200 RPM, that 1900 RPM sample you provided will get used to represent 9200 RPMs. RealFlight will adjust its sample rate to make it match what it should sound like at 9200. That means the pitch will be higher, and it will be played faster. Much, much higher and much, much faster in this case, because that gap is so large.

If you're recording an engine that really only goes to 1900 RPM, then that's one thing. But you should record the entire operating range of whatever engine you're using, so if it goes higher, don't stop halfway there.

Having said all that: you're certainly welcome to try it out and learn for yourself why it doesn't work well. You won't break anything, and that's probably better than me trying to describe it, anyway. :)
 

gibson

New member
Well I guess what I'm asking is that, if I can name the audio file anything I want, can't I record the engine at 1900 (it's max) rpm and NAME the file 4000 rpm or whatever the maximum rpm is, that way, at full power, the sound would be accurate, I would think
 

Ryan Douglas

Administrator
Staff member
Yes, you could do that in order to use the sound profile you created with a motor that revs higher, and RealFlight would avoid having to drastically adjust the sample rate for that top sample, which is what causes the problems I described above.

It would sound "right" quality-wise. Like you said, at full throttle (say, 10000 RPM), you would hear the sound of the motor you recorded at its full throttle (1900 RPM). You would essentially be stretching the sound profile to cover a wider RPM range.

It would not sound correct, however. One thing you're not taking into account is that the sound of two different motors operating at the same RPM have different character--sometimes very different--but they actually have the same pitch. So the motor running at 10000 RPM would sound like it was only turning 1900 RPM.

You might be happy enough with the results. You should try it out and see what you think. Just be aware of the compromise you're making.
 

Maj. Numbskully

Well-known member
Maybe its Time to bring Back both the Resonance and Pitch Sound Settings That where in 4.0 and earlier that were removed from the Editor with 4.5's Release

Remember those ?
No pun intended but It sounds to me that they would now be very useful ??
 
Last edited:
Top